Hi guys,

I'm looking for a method to reliably indicate a cracked shaft. For external components, we normally use dye penetrant testing. But, I'm looking for a method to check the shaft inside the core of the rotor. Is there anything out there that can help?

Thanks, Sean

Mag particle and dye penetrant and normal UT of course rely on having access to the cirfumference of the shaft in the area of inspection.

Supposedly you may see some vibration symtpoms for cracked shaft: appearance of 2x, change in phase. I doubt you would call these reliable means of detection.

For RCP pump shafts, Westinghouse and Areva have a phased-array UT system that mounts on the end of the shaft and looks "through" the length of the shaft to detect cracks. It requires a lot of analysis to understand and interepret those results (including detection threshholds). You could contact them, but I don't think it will be cheap.

I don't think it is a typical failure mode to have cracking on shaft inside the core. Out of curiousity, why do you want to monitor for that?

In the minimg industry (and others I'm sure) we used UT exclusively to determine cracks in shafts. Wasn't really more than what a consultant would charge now (~ $100.00/Hour + exp.). The technology is sound, but it does take a skilled analyst to determine the measurements of crack (length, depth, breadth).
I have seen it work too many times to discount it's worth.
I would say it is like anything, what is the criticality of the shaft?
You can learn to use the technology.
I used to have a Krautkramer ?, and took training from the folks I used to use when I was at the mines. I learned to read them, but didn't stay up with training. The box at the time was around $8k with transducers.
D

Dave - I'm not very familiar with what are the capabilities of routine UT testing that can be done cheaply and easily. But I know the version offered by Westinghouse and Areva for this particular application was not routine and not in the same ballpark, neighborhood, zip-code or county as the prices you mentioned. To do a valid test at our plant would have required constructing or obtaining an identical shaft for calibration purposes. Perhaps the difference is in the size of crack which is reliably detected and the distance (somewhere around 8 feet I think)...I'm not sure.

So in summary, I guess I learned from your post that there is something that can be done a lot cheaper and easier than I described. I just don't know enough to know the differnce in what each test would do.

I was also under the impression that routine UT testing was done from the circumference, not the end of the shaft. Was that an incorrect assumption?

This message has been edited. Last edited by: electricpete,

I have no opinion on effectiveness of this method, could be very good one, but why not use a bump test on a shaft in two orthogonal directions?

Does someone have any success with it?

Hi Sean

LTNS!
I have conducted impact (force/response) tests with force hammer at one end and accel at the other end axial direction. The coherence plots from 0-500 Hz worked nicely. I think the crack would have to pretty far along for this method to work. Also helps if you have a reference good shaft. Loose iron will show up as non-linearities (dips in coherence) as well.

EPete,

We routinely checked shafts from the end. I used to check conveyor drums myself for cracks from the end. I used what is called the pitch-catch method, which utilized two transducers, one sending the signal, and the other listening on the other side. Transducers play an important part in this tech. These folks had a patent on a couple different eddy current type for surface imperfections.
The folks I hired are out of Ill, Ultron Inc. They are a little different in that they develop a lot of the testing methods themselves. The impossible is a challenge they like.
http://www.ultroninc.com/index.htm
They checked dragline walking shafts for me that were in the neighborhood of 14 feet long, and about 4 foot diameter. We didn't have spare shafts to use for calibration. However, they would spend some time on the first one setting the instrument up to ensure they had enough db to get through the shaft. They checked for both internal and surface cracking. Depending on what type transducer they were using (angle of incidence) was what and where they checked. I have seen it work too many times (I was one of the ones who would cut the shaft into next to where they said the crack was, and then chuck it up in a lathe and turn it till I found what they said I'd find) . Length, width, and depth. It was spooky how good a couple of them were. They never missed a call on me, although a couple of them missed dimensions on a regular basis. I don't think they could see the screen divisions , but I paid them well. They did me a good job, and responded in the middle of the night and/or weekends if they needed to. If your in maintenance, and responsible for equipment, you need a service company that will do that.
They also checked castings for me, and X-rayed my hoist and pendant lines and sockets.

Dave

This message has been edited. Last edited by: RRS_Dave,